feat(ai): use openAI provider for suggested fix #69161
Conversation
github-actions
bot
added
the
Scope: Backend
| @@ -264,28 +250,18 @@ def describe_event_for_ai(event, model): | |||
| return data | |||
|
|
|||
|
|
|||
| def suggest_fix(event_data, model=settings.SENTRY_AI_SUGGESTED_FIX_MODEL, stream=False): | |||
| def suggest_fix(event_data, model=settings.SENTRY_AI_SUGGESTED_FIX_MODEL): | |||
| """Runs an OpenAI request to suggest a fix.""" | |||
| prompt = PROMPT.replace("___FUN_PROMPT___", random.choice(FUN_PROMPT_CHOICES)) | |||
| event_info = describe_event_for_ai(event_data, model=model) | |||
There was a problem hiding this comment.
Do we want to use the model configured in the usecase config instead of settings.SENTRY_AI_SUGGESTED_FIX_MODEL now?
There was a problem hiding this comment.
ah good point, this parameter shouldn't be used anymore
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #69161 +/- ##
===========================================
- Coverage 81.44% 79.69% -1.76%
===========================================
Files 2843 6433 +3590
Lines 176165 285413 +109248
Branches 31430 49140 +17710
===========================================
+ Hits 143479 227450 +83971
- Misses 32683 57595 +24912
- Partials 3 368 +365
|
JoshFerge
force-pushed
the
jferg/llmservice--2
branch
from
95f47ca
to
64ced53
Compare
|
This pull request has gone three weeks without activity. In another week, I will close it. But! If you comment or otherwise update it, I will reset the clock, and if you add the label "A weed is but an unloved flower." ― Ella Wheeler Wilcox 🥀 |
instead of using openAI directly, use the new LLM usecase abstractions. Note we'll want to wait to merge this until we have documentation for self hosted, as they'll have to configure providers.